Universities are have become dependent on digital information infrastructures, bringing them into the domain of what Nick Srnicek describes as platform capitalism. Learning Management Systems, MOOCs, teleconferencing facilities, database management systems, and a host of other networks are constructed and often contracted from private companies, such as Google and Microsoft. When academics at Monash University and a number of other Australian universities committed to industrial action over the past few months, it occurred to me: striking academics might bring management to the bargaining table, but a striking IT department would bring them to their knees. (Fortunately for senior managers, IT services can be sub-contracted from other firms, so that any “strike” action would be a mere failure to deliver services with no benefit for the “strikers”.)
What I will explore in this post is not the impact of formally contracted ICT infrastructure and services, but rather those forms of data collection and management that seem to have a life of their own. You might think here of the proliferation of GoogleScholar as a search engine for scholarly work, or Academia.Edu/ResearchGate as public profiling platforms, or the even more bizarre Publons platform that seeks to capitalise on peer-reviewers’ desires for recognition for their work. Where no edict exists to command or encourage scholars to share their personal and professional data, such platforms have thrived.
What I want to suggest is that these practices of searching, uploading and self-archiving are part of a broader pedagogic practice that may be shaping how many early-career academics think about and enact scholarly work. But to do so, some framework is needed to account for the relationship between transforming information systems and user engagement with those systems. I believe that framework may be usefully developed from the concept of ‘data doubles’.
What are Data Doubles?
Before the turn of the 21st century, Kevin Haggerty and Richard Ericson were searching for a means to describe the transformation of surveillance. Whereas the metaphor of “big brother” and Jeremy Bentham’s “panopticon” did a good job of describing the techno-psychology of surveillance in discrete systems (where those being watched are aware of how they are being watched), few attempted to describe how the convergence of several surveillance systems might transform surveillance itself. They argue that surveillance technologies were converging
to the point that we can now speak of an emerging ‘surveillant assemblage’. This assemblage operates by abstracting human bodies from their territorial settings and separating them into a series of discrete flows. These flows are then reassembled into distinct ‘data doubles’ which can be scrutinized and targeted for intervention. (p. 606)
The flows of data that they describe might, in our contemporary context, include not only explicit data that users have offered systems (i.e. personal data, Google searches or publication information) but also meta-data concerning the timing, volume and frequency of data entry. As new modes of data collection join the existing assemblage, the possibilities for the aggregation, combination, filtering and inferences made from data multiply.
In short, data doubles are the ways that data connected to a specific signifier (i.e. your name) are re-assembled to form a meaningful whole. Some data doubles you are likely familiar with already include resumes, GoogleScholar profiles, web search results for persons’ names, performance evaluations. All these data doubles rely on a broad network of data collection points, networks, nodes, infrastructure, and decision-makers (whether human or algorithmic). And because the meaning of such data doubles depends on the data doubles related to other signifiers (i.e. other resumes, other GoogleScholar profiles, other search results and performance evaluations), the final meaning of any data double is always contextual.
The intentional manipulation of data doubles may be described as a ‘speculative’ practice, as the meaning produced through any data double cannot be known until the point of its realisation. Will publishing more peer-reviewed papers increase the perceived value of your ResearchGate profile? It might depend on whether your profile is compared to: scholars at your career stage; of your nationality; of your current era; of your strata of university; etc… As Lisa Adkins has recently described in a different context, speculative practices have the potential to change our relationship to both time and sociality. The value of data doubles and not set in stone, but rather produced through what they might ‘put in motion’ (to use Adkins’ phrase). The double’s value is the promise of some form of input into the conditions of a wager between the academic and an uncertain future.
Managing Data Doubles: Speculatory Prosumption
While some data doubles are clearly in the control of the persons they are seen to represent (you most likely compile your own resume), most seem to approach us, in some form or another. Profiling sites such as ResearchGate and GoogleScholar do not wait for academics to log-on and produce their own input, but generate profiles for them. If you’re depending on your scholarly reputation to sustain your work and career, the onus is then on you to mange the impression that the data double presents to onlookers. George Ritzer and Nathan Jurgenson claim that the work that individuals do on profile-based platforms (and other social media) is both an act of consumption and production. Online impression management produces value for the tech companies that manage these sites – value freely given by users.
What is interesting here is that it is this voluntary labour is enabled by the positioning of the data double within a speculative environment. In other cyber spaces, prosumption may be encouraged by the immediate benefits of productive consumerism: such as Facebook’s social networking and communication functionality or Instagram’s capacity to host large archives of visual data. As Ashlin Lee found in his study of surveillance on Facebook, the perceived benefits of immediate access often outweigh concerns over the potential costs. But the benefits of digital academic profiles are far less certain that the benefits of social media platforms. Data doubles contain a double speculation: that one’s performance through the double is ‘successful’ (i.e. compares well against other people’s doubles) and also that the very existence of that data double means something beyond the platform. In other words: What is a killer Academia.Edu profile actually worth?
The management of data doubles is hence speculatory prosumption. The value that one hopes to add through the labour of profile management is not immediately realised, but rather might be realised at some point in the future. In this sense, it is a kind of hope labour. While there is no obligation to perform speculatory prosumption, the possibility that you are being reckless by not putting your best self forward though these platforms is enough to motivate many to engage beyond any apparent immediate benefit. Labour is effectively extracted from academics and aspiring candidates because “you just never know…”
Early Career Doubling: A Pedagogy?
In labour markets where individualised competition is the norm (i.e. an era of low unionism), digital profile platforms may be seen as one of a range of “coaches” in the arts of competitive self presentation. While William Davies notes that such coaches may include “business gurus, life coaches, ‘leaders’, business lobbyists, motivational speakers and national business representatives”, non-human entities such as digital media platforms may perform similarly disciplinary functions under the right circumstances. Melissa Gregg has already gone some way to describe this disciplinary function as the “athleticism of accomplishment“: self-tracking technologies may be treated as evidence of achievement, to be improved upon and presented as markers of esteem.
This possibly disciplinary function presents some interesting questions for the field of higher education governance research. A vast array of data collection and presentation platforms emphasise the comparability of profiles (chiefly through metricising existing inputs, such as publication counts, or through producing more idiosyncratic metrics, such as ResearchGate’s RG Score). They hence imply an analytic for interpreting the conduct of academics: in this case of GoogleScholar or ResearchGate, as a rivalry between scholars who are ordinarily ranked by the comparison of metrics.
My questions are thus: Might the management of data doubles perform an implicit pedagogical role in the lives of early-career scholars, who are seeking work in saturated job markets? How vital is the input of scholars across different cohorts to the performance of prosumption for digital profiling platforms? And perhaps most pressingly: Has the uptake of data double management influenced the performance of scholarly work? Luka Carfagna’s research suggests that prosumption is not always experienced as labouring for another, but may be felt to be an exchange between persons within a system rather than a quality of the system itself. A deeper ethnography is warranted here.